Skip to content
ALGO BIRDDOG Analytics
Scope live · refreshed 2026-09-27 22:20
Client sign in

Method · Document 01

A rule you can check, not a black box

The proprietary part is small on purpose. What matters is that the rule is measured without look-ahead and reported with its consistency and its sample.

§1 · The signalPrice per square foot, against the neighborhood's own line

Every cohort has a market line: the median price per square foot the neighborhood has been trading at. A listing is a target when it comes to market at or below a fixed percentage under that line. That is the whole rule. It is not a valuation, and it says nothing about condition. It says the seller has priced the house, per foot, where the winning investors historically bought.

The rule was not chosen because it is clever. It was chosen because when we looked back at years of closed sales in North Texas districts, the homes that met it went on to resell within eighteen months far more often than their neighbors — and did so quarter after quarter, not only in a lucky year.

§2 · Point in timeA listing is judged only by what existed before it listed

The line used to flag a listing is computed from sales that closed in the prior quarter. Nothing that happened afterwards is allowed to leak in. This is the difference between a back-test and a story: a story knows how things turned out.

§3 · Fully observed outcomesA flip is counted only when the clock has run

The outcome we measure is a resale within eighteen months. A listing that sold eleven months ago has not had its chance yet, so it is excluded from the tally rather than counted as a miss — or as a hit. No right-censored credit in either direction.

§4 · Walk-forward, by quarterLift is reported with its consistency

The back-test moves forward one quarter at a time and asks, for each quarter, whether the flagged homes beat the neighborhood's baseline resale rate. The proof card a client receives leads with two numbers: the lift, and the count of quarters won out of quarters tested. The noisy quarterly table is printed underneath, not hidden.

§5 · Like with likeSingle-family against single-family; district against district

Investor products compare single-family homes only. The benchmark stays at the district level even when a cohort is a neighborhood cluster, so the lift we advertise is the lift we measured.

§6 · The weekly feedEvery active, ranked; the targets lit

Each week the portal shows every active listing in your cohort. Targets come first. Each row carries the listing firm and agent, the list price and price per foot, how far under the line it sits, the cushion at ask (the distance between the line price and the asking price), days on market, any price cut, and a playbook score from 0 to 100 that grades how closely the house resembles the flips that worked in that district.

What the signal is not

The line price is a screening ceiling, not a rehab underwrite. It tells you where the pattern says the price sits relative to the neighborhood; it does not tell you what the roof costs. Confirm every target with your own numbers and a showing.

Past performance does not guarantee future results. A rule that beat its neighborhood in most quarters lost in some.